[Prev][Next][Index][Thread]

Re: Empty elements (and processing without a DTD)



At 08:24 PM 9/12/96 -0400, Paul Prescod wrote:
>Do people really need conditional markup? Or are they really interested in
>conditionally _displayed_ markup. If the latter, style sheets could be used.
>For instance, here's a graceful mechanism for migrating tables into an
>HTML-like XML instance:
>
><TABLES.DEFINED><TABLE>...</TABLE></TABLES.DEFINED>
><NO.TABLES.DEFINED><PRE>...</PRE>
>
>In a style sheet (in a bizarre syntax similar to my current programming
>language), you could have something like:
>
>TABLES.DEFINED(){
>    if (!FEATURE_SUPPORTED(<TABLE>) )
>       return NULL;
>    else
>        return process_children();
>}
>
>NO.TABLES.DEFINED(){
>    if( FEATURE_SUPPORTED( <TABLE> ) )
>        return NULL;
>    else 
>        return process_children();
>}
>
>Personally, I have never needed to use marked sections in instances for
>anything other than conditional display. I do use them often in DTDs,
>however. Maybe that's an argument against having DTDs be SGML instances.
>Would "XML DTD-instances" themselves have a modifiable DTD that would allow
>me to introduce these sorts of conditionals?
>
> Paul Prescod

Marked sections in instances for conditional display have another
problem: parsers resolve them before any downstream processors see
the instance, so you can't let a user (who's viewing an SGML delivery
form) dynamically switch between two views because only one view is 
left.  This is precisely the reason why DocBook now has "effectivity"
attributes; element and attribute markup can definitely take up the 
slack here.

        Eve